148 The American Geologist. September, 1903. 
modifications, probably migrated from the northeast along- the 
eastern border of the continental nucleus, while the western 
faunas of the same period seem to have come from the north- 
west along the western border of the Devonian continent."* 
As has been stated, the region between Alpena, Michigan, 
on the west side of Lake Huron, and Milwaukee on the west 
side of Lake Michigan, has a mixed fauna, i.e., the Traverse 
fauna is a commingling of Mississippian and Dakota species. 
This mixed Middle Devonic fauna is continued southward as 
far as Louisville and Lebanon, Kentucky, on the west side of 
the Cincinnati island, while immediately on the eastern side of 
this island at Columbus, Ohio, the faunal sequence is more 
decidedly that of New York. The paleontologic work of 
Whitfield on the Devonic beds of central Ohiof shows unmis- 
takably that the succession here is in agreement with western 
Xew York, i.e., the Manlius is followed by the Onondaga, 
Hamilton, Portage and Chemung faunas. In other words, 
the New York Hamilton fauna retains its characters more 
strongly along the shores of Laurentia, Appalachia, and the 
eastern side of the Cincinnati island, while the Traverse Ham- 
ilton fauna followed the eastern shore of the Kankakee penin- 
sula and the western side of the Cincinnati island. To the east 
of the Cincinnati island and west of Appalachia was the Cum- 
berland basin (Williams^), while to the west of the former 
and east of the Kankakee peninsula was another basin which 
may be called the Indiana basin. It is in the latter area that 
the Traverse Hamilton fauna occurs. The Middle Devonic 
faunas of the Mississippian sea belong to the "American 
Province"*: or to the "North Helderberg Sea" of Freeh. || 
Those of the Dakota sea have the general faunal facies of the 
"Eurasiatic Province." 
\Yhatever may be the exact age of the lower formations of 
the Iowa Middle Devonic, it is certain that this area does not 
have the Mississippian Onondaga fauna. Either Iowa with 
* 7owa Geol. Sarv., viii, p. 221. 
t Geol. Surv. Ohio, vii, 1893, pp. 419-464. See also Ortox in same volume, 
pp. 18-26. 
+ The "Cumberland gulf" or basin of Williams should not be confounded 
■with the Cumberland mediterranean of Ulricta and Schucbert. The former 
belongs to the Mississippian province of Devonic time while the latter is an 
arm of the North Atlantic sea beginning with the Ontario and continuing to 
the close of the Oriskanian. 
S Kavskk. Lehrbuch der Geologic sec. ed., 1902. p. 154. 
|| Lethaea Geogrnosticn, 1 Theil, Lethaea Paleozoica, ii, pt. 1, 1S97, paleo- 
graphic map III. 
